Conjoint Analysis

Conjoint Analysis shows respondents hypothetical product "concepts" — each concept is a bundle of attributes (e.g. Brand: X, Price: £29/month, Storage: 100 GB). Respondents choose their preferred concept from each set. By analysing which concepts win across many choice tasks, SurveyKar calculates how much each attribute and each level of that attribute contributes to the likelihood of being chosen.

This tells you what respondents actually value when forced to trade off, not what they say they value when asked directly.

When to use it

Reach for Conjoint Analysis when:

• you are pricing a new product or feature and need to know what respondents will pay
• you want to know which combination of attributes produces the most compelling offer
• you are designing a product with several configurable dimensions and need to know which matter most

How to configure it

Open the question in the builder and expand Settings:

• Attributes — the product dimensions (e.g. Brand, Price, Storage). Add 4–6 attributes.
• Levels — the values each attribute can take (e.g. Price: £9, £19, £29, £49). Add 2–4 levels per attribute.
• Number of tasks — how many choice sets each respondent evaluates (8–12 is standard).
• Concepts per task — how many options appear in each choice set (2–4).
• "None of these" option — allow respondents to reject all concepts in a set.

Tips

Tip — Limit yourself to 4–6 attributes and 2–4 levels each. Every additional attribute increases the number of possible combinations exponentially and makes the exercise cognitively overwhelming.

Tip — Include price as an attribute whenever possible — it is usually the most practically useful output of a conjoint study. Without price, you learn relative feature preferences but cannot directly translate them into pricing decisions.